Volume 123, Issue 6, Pages 1121-1131 (December 2005) Short Telomeres, even in the Presence of Telomerase, Limit Tissue Renewal Capacity  Ling-Yang Hao, Mary Armanios, Margaret A. Strong, Baktiar Karim, David M. Feldser, David Huso, Carol W. Greider  Cell  Volume 123, Issue 6, Pages 1121-1131 (December 2005) DOI: 10.1016/j.cell.2005.11.020 Copyright © 2005 Elsevier Inc. Terms and Conditions

Figure 1 Breeding Scheme and Telomere-Length Analysis (A) Heterozygous generation 1 mTR+/− (HG1) animals were interbred to generate HG2, wt2∗, and mTR−/− KO-G1. KO-G1 animals were intercrossed to generate KO-G2 offspring. Heterozygous animals of the same generation were interbred for five generations. (B) Top panel: Q-FISH telomere-length analysis of wild-type (black, n = 3) and KO-G1 (red, n = 3) animals. The bottom panel shows telomere length of HG1 (gray, n = 4) and HG5 (blue, n = 4) animals. The histogram is truncated at 6600 fluorescence units for clarity; the number of telomeres above this value is plotted at the end on the far right. A Two-sample Wilcoxon rank-sum test showed KO-G1 telomeres are significantly shorter than wild-type (p < 0.0001) and HG1 telomeres are significantly shorter than HG5 telomeres (p < 0.0001). (C) Histogram of the average number of signal free ends (SFE) per metaphase from different animals. Error bars are mean ± SEM. More than seven metaphases were analyzed for each animal and results from multiple animals were pooled for each genotype shown (n = 3 for wild-type, n = 4 for HG1, n = 2 for HG2, n = 3 for HG3, n = 5 for HG4, n = 2 for HG6, and n = 6 for KO-G1). The Q-FISH analysis was scored in a blinded fashion. Cell 2005 123, 1121-1131DOI: (10.1016/j.cell.2005.11.020) Copyright © 2005 Elsevier Inc. Terms and Conditions

Figure 2 Phenotype Analysis of mTR−/− and mTR+/− Animals (A) Kaplan-Meier survival curve of the KO-G1, KO-G2, and wild-type animals shows a statistically significant decrease in survival in the null animals compared to wild-type (p < 0.0001). The median survival for wild-type was 627 days, for KO-G1 it was 242 days, and for the KO-G2 it was 129 days. (B) Kaplan-Meier survival curve of 10 mTR−/− KO-G1 animals that were derived from HG1 cross (green) and 10 mTR−/− KO-G1 animals that were derived from HG3 cross (orange) shows significantly different survival (p = 0.0153). The median survival for KO-G1 from HG1 crosses was 247 days and for the KO-G1 animals from HG3 crosses it was 192 days. (C) H&E section of small intestine. Normal crypts (arrowheads) and villi (arrow) in the wild-type control in the left panel. The middle panel shows an area of severe crypt depletion and villus atrophy. The right panel shows a microadenoma (arrow) with an adjacent area of small intestine with crypt depletion and villus atrophy/loss (arrowheads). The microadenoma had areas with mild dysplasia of intestinal epithelial cells (inset, arrows). All animals were age matched. The slides were read in a blinded fashion. (D) Decreased testes weight and increased aberrant testicular tubules found in the KO and heterozygous animals. Top panel shows the percent aberrant tubules found in each genotype (n = 5 for wild-type, 7 for G1, 2 for G2, and 8 for HG3). Bottom panel shows the testes weight for each animal (n = 9 for wild-type, 4 for G1, 2 for G2, and 9 for HG3). Error bars are mean ± SEM. ## indicates p value of less than 0.05 using Student's t test. (E) GI tract defect in mTR+/− HG5 animal observed in the small intestine (top panels) and large intestine (bottom panels). The wild-type controls are shown to the left. The top right shows crypt proliferation (arrow) with adjacent areas of crypt depletion in the small intestine. The bottom left panel shows normal colon (left) with colonic crypts (arrowheads) and flat intercrypt table surface epithelium. The bottom right image shows an irregular intercrypt table epithelial surface and hyperplasia of colonic crypts (arrow) in the HG5 mouse. Scale bars represent 50 μm. (F) Quantitation of the percent of villous atrophy from large and small intestines from wild-type (n = 9), HG3-HG5 (n = 10), and KO-G1 (n = 7) animals from H&E sections shows more severe atrophy in the heterozygous as well as the KO-G1 animals compared to wild-type controls. Cell 2005 123, 1121-1131DOI: (10.1016/j.cell.2005.11.020) Copyright © 2005 Elsevier Inc. Terms and Conditions

Figure 3 Hematological Defects (A) White blood cell, platelet, and red blood cell counts were analyzed in HG1, HG2, HG3, HG4, HG5, and HG6 mice. A total of 13 animals were analyzed. The counts in each mouse are shown as a dot and the decrease with each generation is shown as a regression line (left panels). The p value for white blood cells, platelet, and red blood cells are 0.1585, 0.0012, and 0.487, respectively. The panels on the right show the blood cell counts plotted relative to the mean telomere length of the individual animal. Due to animal-to-animal variation there is a better correlation of short telomeres with reduced blood counts than the correlation with heterozygous generation. The p value for white blood cells, platelet, and red blood cells are 0.0473, 0.0003, and 0.3041, respectively. (B) The peripheral white blood cell count from wild-type (wt) and HG4 and mTR−/− G1 (KO) were analyzed before 5-FU injection and 6 and 13 days after injection. Error bars are mean ± SEM. Cell 2005 123, 1121-1131DOI: (10.1016/j.cell.2005.11.020) Copyright © 2005 Elsevier Inc. Terms and Conditions

Figure 4 Bone Marrow Transplantation CAST/EiJ recipient mice were irradiated and injected with marrow from HG5, KO-G1, and wild-type animals. Wild-type bone marrow fully rescued viability in recipients for more than 100 days (n = 4) whereas the knockout donor cells was deficient in rescuing viability (n = 4, p = 0.0067 when compared to wild-type using Log-rank test). The heterozygous donor cells partially rescued (n = 7, p = 0.1041 when compared to wild-type using Log-rank test). The unreconstituted control animals (labeled as None) died within 3 weeks after irradiation (n = 5). Cell 2005 123, 1121-1131DOI: (10.1016/j.cell.2005.11.020) Copyright © 2005 Elsevier Inc. Terms and Conditions

Figure 5 Occult Genetic Disease in wt∗ (A) The breeding scheme to derive wt5∗ animals. (B) Telomere-length analysis using Q-FISH in three mouse families. Top panel: wild-type telomere lengths are from animals obtained directly from The Jackson Laboratory. The HG5, wt5∗, and KO are three littermates from a HG4 cross. Middle panel: the HG4 and wt4∗ littermates from a HG3 cross are shown. Bottom panel: the HG5, wt5∗, and KO are three littermates from a HG4 cross. (C) H&E section of control wt (upper) and wt4∗ (lower) testes showing aberrant tubules with degeneration/hypoplasia of germ cell epithelium (arrows). GC = germ cell epithelium; ST = seminiferous tubule. (D) Left panel: the percent aberrant tubules seen in histological sections from wt∗ animals (n = 4) and wt (n = 5). Right panel: the testes weight of wt∗ (n = 9) and wt control animals (n = 6). Error bars are mean ± SEM. ## indicates a p value of less than 0.05 using Student's t test. The analysis was scored in a blinded fashion. Cell 2005 123, 1121-1131DOI: (10.1016/j.cell.2005.11.020) Copyright © 2005 Elsevier Inc. Terms and Conditions